摘要 |
An image processing unit detects a region that corresponds with a subject from a captured image sequence in which a camera and the subject move, based on the three-dimensional shape and motion of the subject. Regions included in captured images are extracted, correspondence is established between the extracted regions, the shape of the corresponding region is estimated by using three-dimensional positions of feature points in the corresponding region, rigid motion of the corresponding region is estimated by calculating motion of each feature point, and region integration or separation is performed based on the estimated rigid motion, whereby the amount of image feature miscorrespondence can be reduced. |